The crystal structures of naturally occurring and assynthesized zeolitic aluminosilicates are composed of AlO.sub.4.sup.- and SiO.sub.4 tetrahedra which are cross-linked by the sharing of oxygen atoms. The electrovalence of each tetrahedron containing an aluminum atom is balanced by association with a cation. Most commonly, this cation is a metal cation such as Na.sup.+ or K.sup.+ but organic species such as quaternary ammonium ions are also employed in zeolite synthesis and, in some instances, appear as cations in the synthesized product zeolite. In general, the metal cations are, to a considerable extent at least, replaceable with other cations including H.sup.+ and NH.sub.4.sup.+. In many instances, the organic cation species are too large to pass through the pore system of the zeolite and, hence, cannot be directly replaced by ion exchange techniques. Thermal treatments can reduce these organic cations to H.sup.+ or NH.sub.4.sup.+ cations which can be directly ion-exchanged. Thermal treatment of the H.sup.+ or NH.sub.4.sup.+ cationic forms of the zeolites can result in the substantial removal of these cations from their normal association with the AZlO.sub.4.sup.- tetrahedra, thereby creating an electrovalent imbalance in the zeolite structure which must be accompanied by structural rearrangements to restore the electrovalent balance. Commonly when AlO.sub.4.sup.- tetrahedra constitute about 40 % or more of the total framework tetrahedra, the necessary structural rearrangements cannot be accommodated and the crystal structure collapses. In more siliceous zeolites, the structural integrity is substantially maintained, but the resulting "decationized" form has certain significantly different properties from its fully cationized precursor.
The relative instability of aluminum in zeolites, particularly in the non-metallic cationic or the decationized form, is well recognized in the art. For example, in U.S. Pat. No. 3,640,681, issued to P. E. Pickert on Feb. 3, 1972, there is disclosed a process for extracting framework aluminum from zeolites which involves dehydroxylating a partially cation-deficient form of the zeolite and then contacting it with acetylacetone or a metal derivative thereof to chelate and solubilize aluminum atoms. Ethylenediaminetetraacetic acid has been proposed as an extractant for extracting aluminum from a zeolite framework in a process which is, in some respects, similar to the Pickert process. It is also known that calcining the H.sup.+ or NH.sub.4.sup.+ cation forms of zeolites such as zeolite Y in an environment of water vapor, either extraneous or derived from dehydroxylation of the zeolite itself, is effective in removing framework aluminum by hydrolysis. Evidence of this phenomenon is set forth in U.S. Pat. No. 3,506,400, issued Apr. 14, 1970 to P. E. Eberly, Jr. et al.; U.S. Pat. No. 3,493,519, issued May 19, 1970 to G. T. Kerr et al.; and U.S. Pat. No. 3,513,108, issued May 19, 1970 to G. T. Kerr. In those instances in which the crystal structure of the product composition is retained after the rigorous hydrothermal treatment involved, infrared analysis indicated the presence of substantial hydroxyl groups exhibiting a stretching frequency in the area of about 3740, 3640 and 3550 cm.sup.-1. The infrared analytical data of U.S. Pat. No. 3,506,400 is especially instructive in this regard. An explanation of the mechanism of the creation of these hydroxyl groups is provided by Kerr et al. in U.S. Pat. No. 3,493,519 wherein the patentees state that the aluminum atoms in the lattice framework of hydrogen zeolites can react with water resulting in the removal of aluminum from the lattice in accordance with the following equation: ##STR1## The aluminum which is removed from its original lattice position is capable of further reaction with cationic hydrogen, according to Kerr et al. to yield aluminum-containing, i.e., hydroxyloaluminum, cations by the equation: ##STR2## It has been suggested that stabilization of NH.sub.4 Y occurs through hydrolysis of sufficient framework aluminum to form stable clusters of these hydroxoaluminum cations within the sodalite cages, thereby holding the zeolite structure together while the framework anneals itself through the migration of some of the framework silicon atoms.
It is alleged in U.S. Pat. No. 3,594,331, issued July 20, 1971 to C. H. Elliott, that fluoride ions in aqueous media, particularly under conditions in which the pH is less than about 7, are quite effective in extracting framework aluminum from zeolite lattices, and, in fact, when the fluoride concentration exceeds about 15 grams active fluoride per 10,000 grams of zeolite, destruction of the crystal lattice by the direct attack on the framework silicon as well as on the framework aluminum can result. A fluoride treatment of this type using from 2 to 22 grams of available fluoride per 10,000 grams of zeolite (anhydrous) in which the fluorine is provided by ammonium fluorosilicate is also described therein. The treatment is carried out for the purpose of improving the thermal stability of the zeolite. It is theorized by the patentee that the fluoride in some manner becomes attached to the constructional alkali metal oxide, thereby reducing the fluxing action of the basic structural Na.sub.2 O which would otherwise result in the collapse of the crystal structure. Such treatment within the constraints of the patent disclosure has no effect on either the overall silicon content of the zeolite product or the silicon content of a unit cell of the zeolite.
Since stability is quite obviously, in part at least, a function of the Al.sub.2 O.sub.3 content of the zeolite framework, it would appear to be advantageous to obtain zeolites having lower proportions of Al.sub.2 O.sub.3 while avoiding the structural changes inherent in framework aluminum extraction. Despite considerable effort in this regard, however, only very modest success has been achieved, and this has applied to a few individual species only.
A process for increasing the SiO.sub.2 /Al.sub.2 O.sub.3 ratio in zeolites is disclosed in copending U.S. Ser. No. 315,853, filed Oct. 28, 1981. The process disclosed therein comprises inserting silicon atoms as SiO.sub.4 tetrahedra into the crystal lattice of an aluminosilicate having a SiO.sub.2 /Al.sub.2 O.sub.3 molar ratio of at least 3 and pore diameters of at least 3 Angstroms with a fluorosilicate salt in an amount of at least 0.0075 moles per 100 grams of the zeolitic aluminosilicate on an anhydrous basis, said fluorosilicate salt being in the form of an aqueous solution having a pH value within the range of 3 to about 7 and brought into contact with the zeolitic aluminosilicate at a rate sufficiently slow to preserve at least 60 percent of the crystallinity of the starting zeolitic aluminosilicate.
The difficulty which is met in preparing titanium-containing molecular sieve compositions is further demonstrated by the failure of European Patent Application No. 82109451.3 (Publication No. 77,522 published Apr. 27, 1983) entitled "Titanium-Containing Zeolites and Method for Their Production as Well as Use of Said Zeolites", to actually prepare titanium-containing molecular sieve compositions. Although the applicants claim the preparation of titano-aluminosilicates having the pentasil structure, it is evident from an analysis of the products of the examples that titanium was not present in the form of a framework tetrahedral oxide. The products of the examples of European patent Application No. 82109451.3 will be discussed in detail in comparative examples hereinafter.
Another reference which deals with titano-aluminosilicates is U.S. Pat. No. 4,410,501 to Taramasso. This reference primarily deals with the preparation of titanium silicates and only in passing does it mention a titanium/aluminum/silicon composition. The patentee presents one example (Example 8) in which it is stated that the addition of aluminum changed the characteristics of the titanium silicate. As will be shown in detail hereinafter, this change in property is owing to the fact that what Taramasso made was ZSM-5 and not a titano-aluminosilicate. In fact Taramasso does not provide any evidence at all to show that titanium was incorporated into the aluminosilicate lattice.